The reliefis composed offour vertical slab-cut boards, reversed
and glued together (Marette 1961). It was reinforced with seven 3 cm thick
limewood boards (five uprights and two crosspieces), apparently glued and
held together by forty screws (Fig. 2). For hanging, two attachments, each
affixed by three screws, were added to the upper crosspiece; the frame was
held in place by four long nails inserted into the sides of the cradle. Two
3 mm wide cracks had run up the whole length of the panel, while two
shorter ones, each about 50 cm long, started upward from the bottom of
the panel. The wood has been heavily eaten by Anobium punctatumworms.
Aprior restoration is indicated by numerous fillings of holes, reconstruc-
tions in places with wood filler, and portions of various sizes reworked
with limewood. Tunnels left by xylophagous larvae, visible on the surface
of the relief, suggest that the panel may once have been painted; the small
remaining amount of ground does not allow us to be more definitive. The
maximum thickness ofthe relief is 10 cm. The cradle, with its two cross-
pieces, created the stresses in the panel that caused the cracking discussed
above. Therefore, the cradle had to be removed.The relief was placed facedown over a thick pad of foam; the cavities in
the wood were then filled with pieces of the same foam cut to size. The
boards were removed one at a time; they were first cut into pieces of vari-
ous sizes according to the thickness ofthe relief and the location of the
cracks. Each piece was then thinned down with a chisel, and the surface
was carefully finished with a scalpel and damp pad, which removed all
traces ofanimal glue. This work uncovered nails and wooden pegs that
had been inserted from the front of the panel to maintain the reconstruc-
tions. The back of the relief being very uneven and the cradle boards quite
flat, the cradle boards had not adhered in all places, and in some areas the
glue was 2–3 mm thick between the relief and the cradle—it is easy to
imagine the stresses these irregularities caused on the surface of the wood.
After cleaning, saw marks became visible on the back, indicating that theTreatment
360 Lebas
Figure 2
Martin Schaffner (attrib.), Annunciation.Back
ofthe reliefbefore restoration.